Dynamic fibrinogen decline rate outperforms static admission values in predicting hematoma expansion in APL-related intracranial hemorrhage: a retrospective cohort study

BackgroundIntracranial hemorrhage (ICH) is the leading cause of early death in acute promyelocytic leukemia (APL). Current risk assessment relies on static coagulation parameters at admission, but APL-associated coagulopathy evolves rapidly. Whether dynamic changes in coagulation markers predict hematoma expansion remains unknown.ObjectiveTo determine whether the fibrinogen decline rate and D-dimer increase rate predict hematoma expansion in APL-related ICH, and to compare their predictive performance with static admission values.MethodsThis single-center retrospective cohort study included 68 APL patients with spontaneous ICH. Hematoma expansion was defined as absolute growth >12.5 mL or relative growth >33% on follow-up CT. To ensure temporal precedence of the predictors relative to the outcome, the fibrinogen decline rate and D-dimer increase rate were calculated over a fixed 24-hour landmark window from the first CT-confirmed ICH (T0) to 24 h (T1). Multivariable models were adjusted for baseline fibrinogen and D-dimer levels. Logistic regression and receiver operating characteristic analyses were performed.ResultsHematoma expansion occurred in 22 patients (32.4%). The fibrinogen decline rate independently predicted expansion in multivariate analysis after adjustment for the baseline fibrinogen level (OR 1.92 per 0.1 g/L/day, 95% CI 1.35–2.74, P = 0.001), while the D-dimer increase rate did not (OR 1.07, 95% CI 0.95–1.21, P = 0.28). The AUC for the fibrinogen decline rate was 0.85, outperforming the static admission fibrinogen level (AUC 0.68) and the absolute fibrinogen nadir (AUC 0.70). The combined model achieved an AUC of 0.94 (bootstrap 95% CI 0.87–0.99). The optimal cut-off for the fibrinogen decline rate was 0.35 g/L/day (sensitivity 77.3%, specificity 80.4%).ConclusionsThe dynamic fibrinogen decline rate independently predicts hematoma expansion in APL-related ICH and outperforms static admission values. Serial coagulation monitoring in the first 24 h may enable earlier identification of high-risk patients. The proposed cut-off of 0.35 g/L/day warrants prospective validation.